I'm wondering just how accurate the dividing lines in the groups of a piece of clothing have to be in order to work properly? I'm currently "cutting" my model into groups so that i can create a .cr2 file in order to get the clothing to conform in poser. I'm following the grouping lines of the figure that the clothing is for. How exact do they have to be to the figure that its made for? I know that the group names have to be precise.

The names have to match the division lines don't have to be exact but it's easier to try and get them as close as possible taking into acccount the geometry of the model. If they are too different from the original the joint bends won't match up. THis can be helped somewhat with JPs but it's easier to try and get them close. mike

cool. that helps a lot. i'm trying to keep it as close as possible but any closer and i'll have to add more verts and edges

It also depends on the poses you expect to use. You can get away with more on a fashion catwalk than a martial arts movie. I've just fiddled with translating P5 clothes to fit Posette-derived figures -- those baggy combat-pants -- and it does look to be the average of the line which matters. A few zig-zags aren't a big problem.

NOT AT ALL. The grouping breaks do not matter AT ALL. Bloodsong's book, "Secrets of Figure Creation with Poser 5", has a wonderful example of this on pages 122 and 123, where she cuts an object in vastly different ways, all with the same joint parameters, and then gives it the same bend. They all move exactly the same. It's good form for the groups to follow the base figure, but they do not need to. The important thing is the joint parameters.

Yes the only things that matter are the group names and the order of hierarchy parenting.

The place where a model bends can be moved around with JP's.

Humm, didn't realize that! Thanks Fyrespiryt and JeffH adds something new to tinker with. Mostly I've just imported "pre-set figures with the set up room to new conforming clothing and I've found that doing it that way it works better if the cuts are realatively close to the originals. mike

"i'm trying to keep it as close as possible but any closer and i'll have to add more verts and edges".....you have it about right there....i mean don't add verts or edges just get it as close as you can with the polys you have ;o).....Steve

I am not sure I agree - but that may be because I am misunderstanding something here.

Are you saying, I can take, my mesh, and section it so that the greater part of the polygons in the mesh are in a group called Head, and then in what might otherwise constitute the foot group, I add in the remaining hiearchy groups in that part of the mesh, and the jp settings in the cr2 will cause the mesh to conform and bend properly as clothing?

If so, I did not know that.

My understanding, is that the body sections need to be an approximate to the grouping in the base figure, but need not be exact. Also, that any group, in order for Poser to be able to handle it correctly with the jp settings, should have a minimal of two rows of connecting polygons.

The comments made here, make it sound almost as if where you section the mesh does not matter. :O) And I think it does, for any of a number of reasons, be it limitations of jp settings, or the nature of the polygons and the way the mesh is constructed, so that the mesh can bend and flex smoothly at the joints being bend, and not tear or break apart. Adjusting the jp settngs alone, can only correct some of that. I have found that I could not elimante certain poke thru problems I was having, via the jp settings alone, but could move the polygons up or down in the effected section, so that where the break in the mesh occurs for the joint is changed, and I could fix the poke thru problem.

The reason for this, is the jp settings have a limited fallout zone, that can be adjusted to some degree, but for clothing to conform, the center points need to match approximately to that of the base figure, and this means that the fallouts will by limited to effecting the surrounding area of mesh in a given way.

Thus, the fallout zones in the collar group are not meant to have an affect the mesh in the forearm group, regardless of how large a sphere you give to the jps in the collar group. The hierarchy is there for a reason, and that is to help Poser know, what what polygons in what groups are to be effected by the jp settings for that group. Since, it does matter where the center points for each body part group are located (at least on conforming clothing), than it also matters what polygons will be affected by what jp settings in what heirarchy group. In this way, the jps settings in the Shoulders tend to effect the Collar group. And the polygons in the shoulder group, tend to also be affected by the ForeArm. What is not meant to happen, is that the jps of the Collar group affect the polygons in the ForeArm group. (Yes, I am sure you can make a cr2 with a jp setting to do that, but this is not the standard heirachy set up for the base figure and clothing.)

So, again, IMHO, I can and do see that it can make a difference in how and where you break polyons in the mesh into different body part sections. The jp settings used have much to say over how well or poorly the polygons in this group flex and distort when bent, however, given that a joint is two jp groups, one interacting with the other, such as shoulder and forearm, then what polygons are in which group, also determines which group works (mostly) on those parts (sections) of the mesh, when the figure is being bent and posed.

Since Poser requires the center point locations for the jp groups to be more or less fixed, then using the jp zones alone, you can only correct so much warping or bending at a joint, based upon, how few or how many polygons are affected in the groups involved. Thus, if you bend the arm, note poke thru at the elbow, and attempt to correct this with just the jp settings alone, you may or may not be able to, depending on how the polygons are sections in the foreArm and shoulder groups. If the bend is causing the polygons in the foreArm group to stretch and tear when flexed, it may be possible to move more polygons into this group by raising where you have the body parts sectioned between those two groups, on that side, so that more polygons are available. Yes, you can do this same thing often via the JP zones alone, but not always.

This is basically the same limitation we encounter when using ghost body parts. I can have a skirt, that is made only of a hip section, and I can set the legs up to work internally using only the jp settings for the thigh groups to cause the skirt to bend and move with the base figures thighs when conformed and posed. I can not, however, add into this hip group, with ghost thighs, ghost shins that are children to the thigh group. Poser jp set up does not allow for the Hip group to extend that far down the hierarchy chain. I could however make the shins a child of the hip group, and thus make them ghost body parts, however they would not work properly as children of the thigh groups in doing so.

Therefore, to me, jp settings alone are not sufficent in correcting how a mesh conforms and poses to the base figure. How the body parts are sectioned, to some degree or another, also factors in - IMO.

I always cut my stuff on the closest existing edge that will match the figure. That is generally enough, with one exception, the DAZ figures that have a very narrow strip of polygons on the back part of the hip will often require a similar, narrow row of polygons in the clothing. That is because of the "hard-edges" that the figure's JP's have incidently. I also try to generally keep the polygon size roughly the same as the figure, too, as taht helps when you have to make JCM morphs and whatnot. I always try and make as few changes in the clothing JPs as possable (compared to the figure), life is just too short, and you will often get into "it helps here, but hurts there"... ;-)

ok. I'm confused again....

i think the short version is your doing it right :o).......Steve

Shadownet seems to be putting it best. The JP's need to be very, very close. The actual groups do not have to be. However, there are limitations on the groups that will generally force you to be similar to the groups of the base figure. It would be one thing if you had a figure with only two body parts. But as most have multiple parts, you can make the abdomen as big as you like....as long as you still have room for the chest....with enough room left for the collars. Grouping is neater if done along edge loops and is greatly similar to the area controlled by the JPs and is also mechanically the area you would like to be moving. Again, not mandatory, but if nothing else making sure the sleeve group is on the sleeve makes it a lot simpler to see what you are doing. I particularly like what Jim says about tweaking the JPs. I've spent weeks myself playing that game, and really, I find coming back to JPs sharing the same centers and having similar movement arms as the base figure does the majority of the work. Spherical fall-off zones are a godsend for clothing, though.

"Not close" is relative, too, I found that using the V3 grouping that I already had wouldn't work very well when I refitted the clothing to the GIRL, as her collars are rather different (they don't extend into the breast area). So I had to weld the clothing up and recut, loosing the existing morphs. On the other hand, I don't cut her breast parts out in the clothing, following the "last group in the chain can be a ghost" rule, in this case the cutting is so far off it doesn't exist. ;- As Steve said, what I'm trying to say is your doing it right.

I just rigged a pair of shorts with little "flags" for the buttocks, meaning I could have ghost thighs that actually connected to buttocks then to hips. It seemed to work -- the problems look do to my own incompetent setting of JPs rather than anything in the grouping. (And the simple reality that a pair of shorts does not fold and crease the same way as the girl wearing them). I've been experimenting with leaving buttocks out but building an ERC link into the thighs that looks towards the figure buttocks. Problem is that the centers of movement are not the same. Long skirts are just a chore.

ok, well. Nice to hear that I'm doing it right, since I've spent so long doing this so far, I'd be extremely piturbed to find I'm doing it wrong and have wasted all these hours. I'm using wings for modeling. I've never messed with grouping in wings yet. I don't know if its the same principle across all/most modeling programs. But, after everything is grouped, and the object is exported, does the grouping cause the idem to be cut into different items? So if I imported the object back into wings each group would now be a seperate item? Is that what happens in other programs? I haven't done this yet. I guess the easiest way to find out is to do it, once I finish the grouping which is almost done. The other question, and more important, is... All the tutorials I've read on creating a .cr2 file for new clothing say to modify the .cr2 of a similar piece of clothing. Now, if i do that, will that prevent me from being able to distribute my new clothing, since I didn't create the .cr2 for it but just modified an existing one? If so, then I don't want to use an existing .cr2. What do I do to create a brand new one from scratch? is there a program available for doing this?

Unless CR2s have morphs or some other imbedded data you are "borrowing" they are just a collection of settings, and they are freely redistributable.

If you really feel queasy about "borrowing" a cr2, use one of the freely-distributable ones from DAZ (like the Vicky LE). Strip out all the morphs and remove the fingers and eyes (ouch!) and save it as a base cr2 for your further use.